An Explainable Intellectual Property Protection Method for Deep Neural Networks Based on Intrinsic Features | IEEE Journals & Magazine | IEEE Xplore